Abstract

The whole literature on the so-called "dimensional effects of the amodal completion" (the shrinkage of a partially occluded figure and the expansion of the modally visible parts of the same figure) is critically reviewed. It is concluded that the causal link between these effects and the amodal completion has never been convincingly demonstrated, and that the stratification in phenomenally superimposed figures accompanies the two dimensional effects, but is irrelevant to their generation. It is maintained that the two dimensional effects are optical-geometrical illusions independent of the figure-ground distinction; phenomenal shrinkage, in particular, is the sum of at least two illusions: (1) the shrinkage of spaces divided in a few parts, and (2) the shrinkage of empty spaces. It is shown, even by new experimental data, that a partially occluded figure can become wider, rather than shrink.